SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2409.00587 · 2024

FLUX that Plays Music

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 9 functions out of this paper's own repositories and ran 6 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
feizc/fluxmusic canonical 6 of 9
FunctionStatusWhere it lives
get_padding Ran feizc/fluxmusic/audioldm2/hifigan/models.py
pointer only (licence: NONE) · get_code("a26f85d7c72ef39a")
get_param_num Ran feizc/fluxmusic/audioldm2/utilities/model.py
pointer only (licence: NONE) · get_code("f1cba67eb6d6e4ad")
prepare Ran feizc/fluxmusic/sample.py
pointer only (licence: NOASSERTION) · get_code("fb7ac27a448677a1")
rope Ran feizc/fluxmusic/modules/layers.py
pointer only (licence: NONE) · get_code("17724d4203fa5863")
swish Ran feizc/fluxmusic/modules/autoencoder.py
pointer only (licence: NONE) · get_code("5fb036462ae82f9f")
torch_version_orig_mod_remove Ran feizc/fluxmusic/audioldm2/utilities/model.py
pointer only (licence: NOASSERTION) · get_code("4dab43e0c6c8c170")
apply_rope Not yet run feizc/fluxmusic/modules/layers.py
pointer only (licence: NONE) · get_code("3444bbcc9065f707")
attention Not yet run feizc/fluxmusic/modules/layers.py
pointer only (licence: NONE) · get_code("b17d74354ecaf7fe")
get_available_checkpoint_keys Not yet run feizc/fluxmusic/audioldm2/utilities/model.py
pointer only (licence: NOASSERTION) · get_code("89bb9be7815a9834")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

This paper explores a simple extension of diffusion-based rectified flow Transformers for text-to-music generation, termed as FluxMusic. Generally, along with design in advanced Flux\footnote{https://github.com/black-forest-labs/flux} model, we transfers it into a latent VAE space of mel-spectrum. It involves first applying a sequence of independent attention to the double text-music stream, followed by a stacked single music stream for denoised patch prediction. We employ multiple pre-trained text encoders to sufficiently capture caption semantic information as well as inference flexibility. In between, coarse textual information, in conjunction with time step embeddings, is utilized in a modulation mechanism, while fine-grained textual details are concatenated with the music patch sequence as inputs. Through an in-depth study, we demonstrate that rectified flow training with an optimized architecture significantly outperforms established diffusion methods for the text-to-music task, as evidenced by various automatic metrics and human preference evaluations. Our experimental data, code, and model weights are made publicly available at: \url{https://github.com/feizc/FluxMusic}.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2409.00587")
get_code_for_paper("2409.00587")
have("2409.00587")

Connect an agent — have() is free.